What are Quadrants? Applying Learning Theory to Dog Training

There are only two main principles for dog-friendly training.  Give something to get more of a behavior you want. Take something away to stop a behavior you don’t like.

Positive Reinforcement (+R): If you want your dog to repeat a behavior more frequently, reward that behavior in some way.

Negative Punishment (-P): If you want your dog to repeat a behavior less frequently, remove any reward or perceived award for the behavior. This should happen rarely – focus on reinforcement.

Access to anything interesting is a reinforcer
Access to anything interesting is a reinforcer

Think of positive and negative in the addition/subtraction sense.  The counterparts to +R and -P are negative reinforcement (take away an aversive – something painful or unpleasant to the dog – as a reward) and positive punishment (present the dog with something painful or unpleasant for doing something you do not like).

“Positive training” usually uses positive reinforcement and negative punishment exclusively, or very rarely uses the other two techniques, and even then with aversives that are not painful, just unpleasant (like time-outs or startling noises).  I do not consider slip chains (choke chains) or prong collars to be part of “positive training,” although some other trainers think of that as “balanced training.” What they don’t know is that a combination of aversive corrections and positive reinforcements has been proven to be the least effective way to teach, and it’s definitely not necessary.

Why go positive?  The reason that many trainers prefer to use +R and -P instead of the alternatives is the fact that working with rewards is so much more fun, for the human and the dog, than aversives like choke chains.  But does it work?  Definitely!  Positive trainers have successfully trained all sorts of competition dogs, from obedience to agility to tracking.  In fact, clicker trained dogs are usually ready for the obedience ring and agility competitions much faster than dogs trained with leash corrections.  Pet dogs everywhere have also benefited from trainers who use these no-force methods.

Let’s discuss +R and -P in more detail.

Positive Reinforcement:  A reinforcer is anything that your dog likes.  Here is a sample list of reinforcers for one of my dogs:

  • Cooked turkey
  • Natural balance
  • Canned cat food
  • Cheese
  • Toys — most of all, the frisbee
  • Petting by me, but not strangers
  • Belly rub
  • Freedom to go outside
  • Dog Park
  • Freedom to jump on the couch or other furniture to be with me
  • Chase games – I run away, he chases
  • His dinner
  • Putting on the leash (if we’re inside)
  • Praise
  • Running an agility course
  • Chasing squirrels
  • Jumping up to lick my face
  • Chicken bones
  • There are plenty more, but that’s enough of a list for now.  Not all reinforcers for my dog are things that I want him to have access to, so I have to prevent him from getting them.  Not all of the items on this list are always reinforcers.  Petting is not reinforcing, even if I do it, if it means he has to wait longer to get the frisbee.  It is definitely not reinforcing to have other people pet him.  My other dog, however, will do anything in the world if it means that she will get attention from a human, especially children.  The moral of the story is to find out what motivates your dog, and what you find acceptable.  Sit down and make a list, it really will be useful.

"What've you got there?"Okay, so now that you have a list of items that your dog will work for, use them to reward your dog for doing what you want.  Teach the behavior using a clicker or marker word (see the clicker handout for more information).  Once the dog knows what you want, continue to motivate him to do the behavior on cue by using one of the rewards above.  If your dog knows the cue, you don’t have to reward every single time (praise is still good), but reward occasionally, to keep him hooked on listening to you.

It is important to not use a positive reinforcer as a bribe, but rather as a reward.  The dog doesn’t see the treat until after she is done.  She has to make a leap of faith that there might be a reward in it for her.  The dog performs the behavior and *then* she is presented with a treat.  The exception is when we use food as a lure, to show the dog how to perform the behavior.  As soon as possible, the lure is faded and the treat is then only used to reward.

Negative Punishment: In general, we should set our dogs up for success.  If you don’t think your dog will respond to your cue, don’t give it.  Work at a level where the dog will be successful and gradually add distractions or distance or whatever makes the task more difficult for your dog.  You can also replace a ‘bad’ behavior with a good one.  To keep your dog from jumping, teach her a solid sit-stay, and any time she looks like she might jump, ask for a sit-stay.  Then give her tons of attention.

But sometimes we do need to punish a behavior, because she is doing something that is dangerous to herself or just plain irritating to us humans. But that doesn’t mean we hurt or threaten the dog.  Positive trainers are not permissive, they just don’t like to risk traumatizing their dogs! 

With negative punishment, we take away something that the dog likes: attention, an expected treat, freedom in space (time-outs), etc.  Here’s an example. Let’s say your dog, Fido, barks at you for attention.  What do you do?  Negatively punish his barking.  He barks, you turn around and leave the room, not allowing him to follow.  Or he barks and you look away.  You remove the expected reward (attention) and eventually the behavior goes away.  Better yet, look away and wait for him to sit.  Then reward him by turning back around.  Soon you have a dog who quietly sits for attention! Even better, work on meeting his needs, and the barking will go away.

Sometimes behaviors are self-reinforcing.  That is, they offer your dog rewards that you have nothing to do with.  In that case, you need to make sure the dog does not perform the behavior to get rewarded, by placing her in another room, putting her on leash to prevent jumping, etc.

When you do any sort of training with your dog, always ask yourself, “Is this improving my relationship with my dog or is it harming it?”  If it is harming the relationship, don’t train that way!  That’s why we use positive reinforcement and negative punishment.  Sure, the other methods often (but not always) work, but why risk all of the fallout?

Related Article: Animal Magnetism: Is Your Training Attracting or Repelling?